Part A - Q1 A - Read the ‘Signals.csv’ as DatFrame and import required libraries.
Part A - Q1 B - Check for missing values and print percentage for each attribute.
No missing values in the dataset
Part A - Q1 C - Check for presence of duplicate records in the dataset and impute with appropriate method.
Part A - Q1 D - Visualise distribution of the target variable.
Part A - Q1 E - Share insights from the initial data analysis (at least 2).1. The signal strength peaks at 5. It increases until 5 and then decreases till 7. 2. Parameter 1 is correlated to 8, Parameter 6 is correlated to 7.Part A - Q2 A - Split the data into X & Y.
Part A - Q2 B - Split the data into train & test with 70:30 proportion.
Part A - Q2 C - Print shape of all the 4 variables and verify if train and test data is in sync.
Part A - Q2 D - Normalise the train and test data with appropriate method.
Part A - Q2 E - Transform Labels into format acceptable by Neural Network
Part A - Q3 A - Design a Neural Network to train a classifier. Part A - Q3 B - Train the classifier using previously designed Architecture
Part A - Q3 C - Plot 2 separate visuals. i. Training Loss and Validation Loss ii. Training Accuracy and Validation Accuracy
Part A - Q3 D - Design new architecture/update existing architecture in attempt to improve the performance of the model.
Part A - Q3 E - Plot visuals as in Q3.C and share insights about difference observed in both the models.
Based on the results we can conclude that "Model 2" has a higher validation accuracy and a lower validation loss compared to "Model 1." This suggests that "Model 2" is performing better on the dataset.Part B - Q1 A - Read the .h5 file and assign to a variable. Part B - Q1 B - Print all the keys from the .h5 file.
Part B - Q1 C - Split the data into X_train, X_test, Y_train, Y_test
The dataset already seems to be split into 70,30 ratio.Part B - Q2 A - Print shape of all the 4 data split into x, y, train, test to verify if x & y is in sync.
Part B - Q2 B - Visualise first 10 images in train data and print its corresponding labels.
Part B - Q2 C -Reshape all the images with appropriate shape update the data in same variable. Part B - Q2 D - Normalise the images i.e. Normalise the pixel values.
Part B - Q2 E - Transform Labels into format acceptable by Neural Network
Part B - Q2 F - Print total Number of classes in the Dataset.
Part B - Q3 A - Design a Neural Network to train a classifier. Part B - Q3 B - Train the classifier using previously designed Architecture (Use best suitable parameters).
Part B - Q3 C - Evaluate performance of the model with appropriate metrics.
Part B - Q3 D - Plot the training loss, validation loss vs number of epochs and training accuracy, validation accuracy vs number of epochs plot and write your observations on the same.
Conclusion: Training vs. Validation Loss The validation loss is slightly lower than the training loss. This indicates that the model might still be improving, or it could be a result of variations in the validation dataset. However, the difference is not significant, suggesting that the model is not overfitting. Training vs. Validation Accuracy The final validation accuracy is approximately 78.20%, and the training accuracy is 75.84%. The model's performance on the validation data is in line with its training performance, which is a good sign. It suggests that the model generalizes well to unseen data. Overall Performance The model is performing reasonably well, with a final validation accuracy of 78.20%. However, the interpretation of this accuracy depends on the specific image recognition task and the dataset.